The paper is available at: LNCS webpage
This is a repository for the code to generate and use Gauss-Hermite approximation for EHVI calculations.
The required Python libraries are:
To install the exact EHVI code, you need to go to "EHVI/EHVI_2D" and "EHVI/EHVI_3D" directories in turn, and issue the following command.
$ g++ -Ofast -o EHVI -std=c++0x *.cc
An illustration on how to use the code is provided in the "notebooks/Illustration of running experiments.ipynb". You would need Jupyter notebook to run that.
The data from the experiments ran for the paper is avaialble in the "data/experiments/" directory. For the meaning of the headers of that file, please refere to the illustrative notebook, in particular, the last paragraph where we introduced the keys to the labels.
Any queries, suggestions, or comments should be directed to Dr Alma Rahat.